This study classified groundwater level variation types in rural areas by analyzing long-term groundwater level time series and hydrogeological variables from 157 monitoring wells. Groundwater level trends and variability were quantified using the Mann-Kendall test and Sen’s slope estimator, and a multivariate dataset including hydraulic conductivity, elevation, depth, hydrogeological unit (HGU), and pumping rates was used to perform hierarchical cluster analysis and principal component analysis (PCA). The cluster analysis identified four distinct groundwater level variation types-persistent decline, stable increase, low variability, and high variability in highly permeable aquifers-reflecting differences in groundwater response to geological, topographic, climatic, and anthropogenic factors. PCA showed that PC1 was dominated by hydraulic conductivity, while PC2 was primarily governed by groundwater level trend with secondary hydrogeologic influences. These components effectively summarized major groundwater variation patterns. Visualization of the four clusters in the PC1-PC2 space showed clear separation among groups, demonstrating that PCA provides a concise and robust framework for distinguishing groundwater level variation patterns in rural monitoring networks.

Keywords: Principal component analysis (PCA), Hierarchical cluster analysis, Groundwater level variation type, Groundwater monitoring wells, Hydrogeological unit (HGU)